This files is hosted for you with Grady Hull’s permission by
www.BrewCrAzY.com
The Podcast for those of us who Brew our Own Damn BEER!
Olive Oil Addition to Yeast as an
Alternative to Wort Aeration
Grady Hull
New Belgium Brewing Company, Fort Collins, CO USA
9/26/05
Submitted to Professors Graham Stewart and Brian Eaton
in Partial Requirement of
Postgraduate Diploma/MSc in Brewing and Distilling
__________________________________________________________________________________
Table of Contents
_____________________________________________________________________
Acknowledgments
4
List of Tables and Figures
5
Abstract
7
Introduction
7
Background
8
Wort Aeration
8
Yeast Aeration
9
Addition of Unsaturated Fatty Acids
10
Materials and Methods
11
Fermentations
11
Aeration and Olive Oil Addition
12
Sensory Analysis
13
Gas Chromatography
13
Analytical Analysis
13
2
Results
14
Trial One
14
Trial Two
16
Trial Three
19
Trial Four
23
Discussion
26
Results
26
Oxidative Effects of Wort Aeration
27
Ester Production and Growth
27
Suggestions for Future Work
27
Conclusion
29
Literature Cited
30
Appendix 1 Fusel Oils by Headspace GC SOP
32
3
______________________________________________________________
Acknowledgements
______________________________________________________________
First of all I would like to thank the New Belgium Brewing Company for allowing me
to conduct this research at their brewery.
Thank-you to Peter Bouckeart for mentoring me throughout this project.
I would like to thank Dr. Ir Guy Derdelinckx for his ideas and direction.
Thank-you to Lauren Salazar for running the flavor panels and to Todd Bennet for
performing the GC analysis.
I would also like to thank Jason Trujillo, Simon Ahlgren, Katie McGivney and the rest
of the New Belgium quality assurance department for all of their analytical work.
4
______________________________________________________________
List of Tables and Figures
______________________________________________________________
Table
Page
1.
Results for the first olive oil test fermentation vs. the control.
14
2.
Results for the second olive oil test fermentation vs. the control.
17
3. Results for the third olive oil test fermentation vs. the control.
19
Figure
1. Gas Chromatography results for the first olive oil test vs. the control.
15
2. Sensory results for the first olive oil test vs. the control.
16
3. Gas Chromatography results for the second olive oil test vs. the control. 17
4. Sensory results for the second olive oil test vs. the control.
18
5. Gas Chromatography results for the third olive oil test vs. the control at
20
fresh.
6. Gas Chromatography results for the third olive oil test after 3 weeks of
21
being stored at room temperature.
7. Sensory results for the third olive oil test vs. the control.
22
8. Sensory results for the third olive oil test after 3 weeks room temp.
23
9. Gas Chromatography results for the fourth olive oil test vs. the control.
24
5
10. Sensory results for the fourth olive oil test vs. control.
25
11. Sensory results for the fourth olive oil test after 3 weeks warm storage.
25
6
Abstract
To extend the flavor stability of their beers, many breweries are researching ways of
reducing oxygen ingress throughout the brewing process. However, the practice of
aerating the wort prior to fermentation is almost universal in the brewing industry
because oxygen is necessary for yeast health and growth. Recent studies have
shown that alternative methods to traditional wort aeration such as aeration of the
yeast prior to pitching or the addition of the unsaturated fatty acid linoleic acid can
yield fermentation characteristics similar to wort aeration. It has also been shown
that using these alternative methods instead of aerating the wort can reduce
oxidation potential. This paper reports the findings of a series of full-scale production
tests that were conducted in an operating brewery to evaluate the effects of another
type of yeast treatment. By mixing olive oil into the yeast, during storage, instead of
aerating the wort, fermentations can be achieved with only minor increase in
fermentation time. The beers produced from these fermentations were comparable
in flavor and foam retention to beers produced by traditional wort aeration. The ester
profile of the beers produced using olive oil addition was significantly higher than the
controls and the flavor stability of these beers was significantly improved.
Introduction
In modern breweries a great deal of care is taken to ensure that oxygen is excluded
from coming into contact with the beer or unfermented substrate (wort) throughout
the brewing process. The one exception to this practice is that the wort is typically
injected with air or pure oxygen (aerated) prior to fermentation. Although it is
universally accepted in the brewing industry that oxygen contact with beer or wort
leads to staling of the product, wort aeration is considered necessary because
without it the yeast will not be healthy enough to properly ferment the wort (Kunze,
1999).
7
The reason the yeast needs oxygen for a proper fermentation is because it needs to
synthesize sterols and unsaturated fatty acids (UFAs) for its cell walls. Yeast is
typically collected from a fermenter at the end of fermentation and stored in a storage
tank. The yeast is then re-used in a subsequent fermentation. The physiological
condition of the yeast is usually poor at the time of re-use due to depletion of sterols
and unsaturated fatty acids from the previous fermentation (Moonjai, 2003a).
Therefore oxygen is typically added to the wort at the start of every fermentation.
One interesting alternative to aerating the wort is to add the UFAs directly to the
yeast during storage. Theoretically the yeast should be able to take up the UFAs
and use them in a subsequent fermentation without the use of oxygen. This should
result in a beer that has better resistance to staling oxidation without adversely
effecting fermentation performance or flavor.
The purpose of this research was to compare the effects of adding olive oil to
storage yeast vs. traditional wort aeration. The theory is that the oleic acid in the
olive oil will provide the UFAs necessary for yeast growth and proper fermentation,
eliminating the need for wort aeration.
In this paper we will look at the results of a study in which full-scale fermentations
were conducted to evaluate the fermentation performance, flavor, fusel profile, and
analytical attributes of beer, which had been fermented with yeast that was treated
with olive oil during storage instead of aerating the wort. The aim is to achieve a
normal fermentation, and flavor profile while improving resistance to oxidative staling.
Background
Wort Aeration
Yeast cells require oxygen to manufacture the sterols and unsaturated fatty acids
needed for cell membrane construction and growth. For this reason oxygen is
typically injected into the wort prior to fermentation to bring the total dissolved oxygen
content of the wort up to about 8-9 ppm. This addition of oxygen to the wort has
traditionally been considered to be essential for yeast growth (Kunze, 1999). If the
oxygen content in the wort is insufficient, the yeast cells will be unable to
8
manufacture the sterols and unsaturated fatty acids necessary for cell membrane
health. As a result the yeast cells do not grow and the loss of membrane integrity
results in cell death (Hough et. al., 1982).
Stored yeast from a previous fermentation is typically deficient in sterols and
unsaturated fatty acids and therefore wort aeration is necessary for yeast health and
growth but over oxygenation of the wort should also be avoided as it can make the
resulting beer more susceptible to oxidation (Anderson et. al., 2000). The amount of
oxygen added to the wort will also effect ester production and subsequently the
flavor profile of the resulting beer. Wort that is not properly aerated will not support
healthy yeast growth and the resulting beer will have an increased ester content and
other flavor defects. If the wort is over oxygenated not only will there be an excess
production of yeast during fermentation, causing a decrease in yield, but ester
synthesis will also be strongly inhibited (Smart, 2003).
During the short period of time that the oxygen is in contact with the wort, harmful
oxidative chemical reactions are taking place that form the precursors of beer staling
compounds. Although the yeast is taking up the oxygen from the wort very quickly,
the oxidative reactions are taking place at a similar rate. As oxygen ingress into the
cold wort is reduced, the formation of these beer staling precursor compounds is also
reduced. If wort aeration could be avoided completely, without adversely effecting
yeast health or fermentation performance, flavor stability should be improved
(Burkert, 2004).
Yeast Aeration
Studies have been conducted to investigate aeration of storage yeast, prior to use,
as a means of avoiding wort aeration. The theory is that the oxygen will be taken up
into the yeast cells during storage, before the yeast is added to the wort. This should
supply the yeast cells with the oxygen they need without oxidizing the wort. There
are, however, some problems with this method of aeration.
The correct amount of oxygen must be dissolved into the yeast in order to achieve
optimal growth during fermentation. Under aeration will obviously lead to low growth
but over aerating can also lead to low growth due to glycogen and trehalose
exhaustion (Fugiwara et. al., 2003). The success of yeast aeration can also be
9
dependent on the yeast strain. Each yeast strain reacts differently to storage
aeration so the aeration schedule must be specifically tailored to each strain
(Depraetere et. al., 2003).
In the initial lag phase of fermentation the yeast cells begin taking up the oxygen
from the wort. Usually at the end of these first few hours all of the oxygen has been
taken up by the cells and during this time the cells rely on glycogen for their
metabolic activity so it is essential that the cells begin fermentation with all of their
glycogen reserves in tact (Hardwick, 1995). For this reason, brewery yeast is
typically stored cold and not aerated during storage to minimize metabolic activity so
that the yeast cells will have enough glycogen and trehalose to remain vital during
this initial lag phase and begin fermentation.
Yeast aeration may not be a practical option for many breweries because oxygen
cannot be added to the yeast too early. Yeast storage is a critical step in yeast
handling and any storage conditions that increase metabolic activity such as
increased temperature, prolonged time of storage, or the presence of oxygen will
cause the yeast to deplete its internal glycogen and trehalose reserves. Also studies
conducted to replace wort aeration with yeast aeration have shown that in order to
replicate the fermentation times from traditional wort aeration it was necessary to
increase the amount of yeast by approximately 30% (Smart, 2000).
Addition of Unsaturated Fatty Acids
Research has been done to investigate adding linoleic acid to wort prior to use but
the result was a change in flavor quality due to an increase in the acetate esters.
Another possible alternative to wort aeration that has been studied is the direct
addition of linoleic acid to storage yeast. Yeast stored in a cold stationary phase
under fermented beer has been shown to take up unsaturated fatty acids.
Additionally, linoleic acid is taken up by the spheroplasts, proving the oil is not just
adsorbed to the cell walls (Moonjai, 2003a).
Research done by Nareerat Moonjai at Leuven University has shown that smallscale, non-stirred fermentations in a wort medium using yeast that had been treated
with linoleic acid instead of wort aeration could produce normal fermentations without
significantly affecting the acetate esters. These findings showed that under typical
10
production yeast storage conditions it should be possible to get fermentation
performance comparable to traditional wort aeration by adding an unsaturated fatty
acid such as linoleic acid to the yeast (Moonjai et. al., 2003b).
For this study olive oil was selected as the unsaturated fatty acid (UFA) over linoleic
acid because it’s much lower cost and wide spread availability make it a more
realistic choice for production fermentations. Also the oleic acid in olive oil is an 18carbon monounsaturated UFA (C18: 1), which is one of the UFA’s that S. cerevisiae
manufacture. Linoleic acid, being a polyunsaturated 18-carbon UFA (C18: 2) is not
naturally produced by yeast (Depraetere et. al., 2003). Also, because these
fermentations were carried out with ale yeast at fermentation temperatures starting at
15.5 °C and rising naturally to 24°C, the improved fluidity of the polyunsaturated
linoleic acid was not required.
Materials and Methods
Fermentations
The same yeast strain was used for all test and control fermentations. It is an ale
yeast, which is pitched during transfer to fermentation. The yeast is then collected
from the bottom of the fermenter at the end of fermentation and stored in yeast
storage vessels to be re-used on a subsequent fermentation. The wort has a starting
density (before fermentation) of 1.057 g/ml and an final density (after fermentation) of
1.010 g/ml. After a fermenter has been cooled the maximum storage times for yeast
were 48 hours in the fermenter cone and 72 hrs in the storage tank. Yeast viabilities
for these tests and controls were above 85%, with the average being 95%. Yeast
pHs were below 4.6. IBU’s (ppm isomerized alpha acids) were around 20.
During the first round of testing the temperature of the wort for both the test and the
control was 15.5 °C at the start of fermentation. The fermentation temperature was
allowed to rise naturally to 20.0 °C during fermentation. After this first round, an
unrelated procedural change was made at the brewery to allow all fermentation
temperatures to rise to 24 °C. Therefore all tests and controls following this first
round had faster fermentation times due to the increased temperature.
11
In the first round of testing a 360 hl batch size was used. The second round was a
720 hl batch, and the third and fourth rounds were 2100 hl. The 2100 hl batches
were kept separate throughout production and were bottled and sold as finished
product without being blended with any of the control beer.
The 360 hl batch was blended in the bright beer tank with 1800 hl of regular
production beer after determining it was safe to do so based on flavor. A keg of
filtered carbonated beer from the test fermenter, which was 100% olive oil test beer,
was collected after post filtration. The beer that was pulled after the filter was tested
for aromatic compounds, foam retention, and flavor. The second 720 hl batch was
blended with 1400 hl of normal production beer in the bright beer tank and again, a
keg of filtered carbonated beer from this tank was pulled after the filter for testing
aromatic compounds, foam and flavor. The 2100 hl test batches were bottled
without blending so that they could be compared directly against packaged control
production beers. These beers were tested for initial flavor quality, aromatic
compounds, foam, and flavor stability.
Aeration and Olive Oil Addition
All controls were aerated in-line, with micro filtered compressed air, in excess of
saturation for the entire duration of the transfer according to the breweries standard
operating procedures. The tests were not aerated. For the test fermentations, olive
oil was added to the yeast in storage tanks five hours prior to use and the amount
added increased with each trial. Due to the variation in yeast slurry thickness the
amount of olive oil used was based on the total number of cells instead of mg / L of
yeast. In the 360 hl batch the olive oil was added to the yeast at a rate of 1 mg / 67
billion cells pitched (15 mg olive oil / L of yeast assuming a count of 1 billion cells /
ml). In the 720 hl trial the concentration was increased to 1 mg / 50 billion cells and
in the 2100 hl trials the concentration was increased again to 1 mg / 25 billion cells.
Aside from the changes previously mentioned with aeration, olive oil addition and
fermentation size, all other aspects of production were carried out identically for both
the tests and the controls.
12
Sensory Analysis
New Belgium Brewing Company’s in-house taste panel conducted all sensory
analyses. All taste panels were comprised of 12-14 internally trained senior level
taste panelists. Qualification for senior level panel requires a minimum of three
years training and 20 hrs of training classes. Panelists tasting ability determined by
weekly flavor attribute testing. Panelists tasting calibration determined by Senstools
General Procrustes Analysis statistical software. Senstools statistical software used
for all ANOVA and Principal Component Analysis. All flavor profile analyses were
performed on a nine-point scale. Sample randomization and data entry organization
performed by Williams complete block method using Compusense sensory software.
Gas Chromatography (GC)
GC was carried out at New Belgium Brewing Company using Perkin Elmer
Autosystem XL GC equipped with a Perkin Elmer Turbomatrix headspace sampler, a
Perkin Elmer PE-Wax column (60m x 0.25mm id x 0.25um film), and a flame
ionization detector (FID). See Appendix 1 for detailed GC procedure.
Paragon Laboratories performed external GC/MS analysis using a Hewlett Packard
5890 GC equipped with an EST Archon auto sampler coupled to an OI 4560 Purge &
Trap sample concentrator, a Restek RTX-624 capillary column (60m x 0.25mm id x
1.4um film), and a Hewlett Packard 5971 mass spectral detector.
Analytical Analysis
A Nibem foam analyzer was used to determine the rate of foam collapse according
to ASBC standardized method. Yeast health and viability were monitored on a
percent viability basis using a hemocyometer and the ASBC standardized method for
the methylene blue staining technique. Fermentation speed was measured from the
start of the first transfer to the time the temperature on the tank was turned down at
the end fermentation. All fermentations were ended when density reduction had
stopped and total Vidicinal Keytone (VDK) levels had reached the brewery’s
specification. VDKs were measured using the ASBC spectrophometric method.
13
Results
Trial One
The results for the first olive oil test fermentation are shown in Table 1. The olive oil
test fermentation was 20% longer than the average aerated fermentation during that
time period but it did attenuate completely. Both had good yeast viability. Foam
was not significantly affected by the test.
Table 1. Results for the first olive oil test fermentation vs. the control.
Sample
Fermentation
Density
Time (hrs)
(g/ml)
Yeast
Nibem Foam
Viability
Collapse (sec)
Olive oil test
140
1.011
94%
272
Control
117
1.011
96%
278
In all of the G.C. figures, the flavor units are the amount of each individual
compound, in parts per million (ppm) as determined by G.C. divided by the
established flavor threshold for that particular brand. Therefore, any compound with
a flavor unit value greater than one is considered to be above flavor threshold for that
brand. The control values given in these GC analysis graphs are an average of all of
the non-test production beer during that time period. The Y error bars given for each
control show the average standard deviation between the control finished product
samples during that time period. The test results do not show standard deviation Y
error bars because they are single test results not averages.
One anomaly in this first test appears to be the absence of dimethyl sulfide (DMS) in
the test sample. This was considered to be an anomaly because it sometimes seen
in normal production beers. Total DMS levels are considered to be more of a
function of the brewhouse boiling system and not related to fermentation.
The ester profile for the test in Figure 1 was higher than that of the control but not out
of the breweries specifications for the brand. Also, in this particular situation, the
14
increase in esters was considered by the brewery to be a positive change because of
the potential to mask staling flavor compounds. We know from historical data that as
this brand ages ethyl hexanoate and iso-amyl acetate both decrease. Ale flavor
compounds such as acetaldehyde, ethyl acetate, isoamyl acetate and ethyl
hexanoate were all more than double the control (Figure 1).
Figure 1. Gas Chromatography results for the first olive oil test vs. the control beers.
GC Analysis Test # 1
2.5
Flavor Units
2
1.5
Olive Oil Test
Control
1
0.5
yl
ac
et
is
at
oa
e
m
yl
al
co
et
ho
hy
l
lh
ex
an
oa
te
an
ol
is
oa
m
is
ob
ut
npr
op
ac
et
al
de
hy
de
D
M
S(
pp
b)
et
hy
la
ce
ta
te
0
The test beer was then put on the brewery’s flavor profile analysis taste panel to see
if the brewery’s flavor panel could detect the increase in esters shown by the GC
results. The sensory results, shown in Figure 2, confirm the increase in
acetaldehyde, ethyl acetate, isoamyl acetate and ethyl hexanoate indicated by G.C.,
however the taste panelists did not perceive the difference as significant so the next
round of testing continued. Note that Figure 2 illustrates that an acceptable flavor
match was achieved between the test and control beers. The comments from that
taste panel indicated an overall preference for the olive oil test.
15
Figure 2. Sensory results for the first olive oil test vs. the control.
Sensory Test # 1
hoppy
4.5
metallic
4
medicinal
3.5
3
diacetyl
2.5
2
fatty acid
1.5
1
0.5
oxidation
0
control- fresh bottle
olive oil test keg
SO2
DMS
ethyl hexanoate
ethyl acetate
acetaldehyde
isoamyl acetate
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
Trial Two
The analytical results for the second test, shown in Table 2, were similar to the first
trial. For this and all subsequent rounds of testing the brewery had made a change
in the maximum fermentation temperature, increasing it from 20°C to 24°C. This
procedural change was unrelated to the olive oil tests. Due to this change all
fermentation times for both tests and controls were decreased. It is interesting to
note that the difference in fermentation time decreased from 20% in the first test to
14% in the second. The rate of olive oil addition for the second test was 33% greater
than in the first.
The difference in the post-fermentation density is significant and it is a concern but it
is not outside of the breweries specification range (1.010 to 1.014 g/ml). The
apparent improvement in foam is also well within the breweries normal variations,
which can fluctuate between 230 and 290 seconds on the Nibem. Again viability was
considered to be normal for both the test and the control (Table 2).
16
Table 2. Results for the second olive oil test fermentation vs. the control.
Sample
Fermentation
Density
Percent
Nibem Foam
Time (hrs)
(g/ml)
Viability
Collapse (sec)
Olive oil test
90
1.014
96%
277
Control
79
1.012
92%
262
The G.C. results for the second round of testing shown in Figure 3 also indicate that
this test was closer to the control than in the first round. Compounds such as
acetaldehyde and ethyl hexanoate are actually within the brewery’s standard
deviation and the remaining fusel compounds such as ethyl acetate and isoamyl
acetate are much closer than they were in the previous round of testing (Figure 3).
Figure 3. Gas Chromatography results for the second olive oil test vs. the control.
GC Analysis Test # 2
1.8
Olive Oil Test
Control
1.6
1.4
1.2
1
0.8
0.6
0.4
0.2
at
e
l
lh
ex
an
o
co
ho
al
et
hy
am
yl
is
o
is
o
am
yl
ac
et
ta
at
e
no
l
p
is
ob
u
npr
o
D
M
S(
pp
b)
et
hy
la
ce
ta
te
ac
et
al
de
h
yd
e
0
17
According to the sensory panel (Figure 4) there was a statistically significant
increase in ethyl hexanoate at a 95% confidence level. While there were no other
attribute changes that could be called significant to that confidence level, the panel
perceived the olive oil test to be higher in other esters such as ethyl acetate and
isoamyl acetate. Other flavor attributes that have been linked to these fruity-ester
compounds such as perceived body, and sweetness were also rated higher in the
test. Despite the increase in ethyl hexanoate shown in Figure 4, the flavor panel
approved both rounds of testing for production release.
Figure 4. Sensory results for the second olive oil test vs. the control.
Sensory Test # 2
metallic
medicinal
diacetyl
fatty acid
oxidation
hoppy
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
control
olive oil test
SO2
DMS
ethyl hexanoate
ethyl acetate
acetaldehyde
isoamyl acetate
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
The sensory, fusel, and analytical testing results for the first two rounds of testing
were all within the breweries specifications so the decision was made to proceed
with the full-scale test. Bringing the test through production as a full scale batch so
that all product handling including packaging are done in the same way as the
control sample was deemed to be the most appropriate way to be sure that noise
from other variables such as the operators, equipment or procedures would be kept
to a minimum.
18
Trial Three
In the first two rounds of testing, esters were increased and higher alcohols were
decreased compared to the control indicating low yeast growth in the olive oil test
(Fig. 1-4). Supporting this assumption were the reduced fermentation speed and
attenuation (Table 1,2). To improve this, the amount of olive oil used in the third
round was increased 100% over the previous round.
As shown in Table 3, again the difference in fermentation time decreased, but this
time only slightly from 14% longer in the second test to 13% longer in the third. Post
fermentation density was much better, compared to the control, in this test than in
the previous test although again both were well within the brewery’s specifications.
Foam, attenuation and yeast viability were all within the normal range.
Table 3. Results for the third olive oil test fermentation vs. the control.
Sample
Fermentation
Density
Percent
Nibem Foam
Time (hrs)
(g/ml)
Viability
Collapse (sec)
pH
Olive Oil Test
94
1.011
90%
263
4.38
Control
83
1.013
97%
269
4.33
The G.C. flavor component results for this third round of testing show a significant
improvement in total esters and higher alcohols (Figure 5). Although there is still an
increase in esters and a decrease in higher alcohols, the difference for most of these
compounds is within the standard deviation of the average control. The only
important beer flavor compound in which the difference exceeds the standard
deviation was isoamyl alcohol.
19
Figure 5. Gas Chromatography results for the third olive oil test vs. the control at
fresh.
GC Analysis Test # 3 Fresh
1.8
1.6
1.4
Flavor Units
1.2
1
Olive Oil Test
Control
0.8
0.6
0.4
0.2
te
oa
an
ex
lh
et
hy
am
yl
al
co
ho
l
e
et
at
ac
is
o
am
yl
is
o
is
ob
ut
an
ol
pr
op
n-
ta
te
la
ce
pb
et
hy
M
S(
p
hy
de
D
ac
et
al
de
-0.2
)
0
The bottles from this run were then stored at room temperature for three weeks and
the beer was re-analyzed by G.C. (Figure 6). As expected acetaldehyde increased
in both the test and control samples although both were well below threshold.
Isoamyl acetate and ethyl hexanoate were both reduced during warm storage
although the test was less effected than the control and these two compounds were
still present at threshold levels after one week warm storage.
20
Figure 6. Gas Chromatography results for the third olive oil test after 3 weeks of
being stored at room temperature.
GC Analysis # 3 after 3 Weeks at Room Temperture
1.8
1.6
Flavor Units
1.4
1.2
1
Olive Oil Test
Control
0.8
0.6
0.4
0.2
te
ol
an
oa
co
h
lh
ex
al
et
hy
am
yl
is
o
is
o
am
yl
ac
bu
et
ta
no
at
e
l
op
is
o
npr
la
ce
ta
t
e
b)
D
M
S(
pp
et
hy
ac
et
al
de
hy
de
0
There are some marker compounds that can be used to quantify oxidative changes.
Compounds such as acetaldehyde will typically increase during storage as a result of
the oxidation of ethanol. Ester compounds such as, ethyl hexanoate, and isoamylacetate will typically decrease with oxidation. Two test bottles and two control
bottles were stored at 30 °C for one week and two test and two controls were stored
cold. All of these bottles were then tested by GC/MS for known oxidation markers for
this brand. The results from the GC/MS analysis were mixed. Previous oxidation
research has shown that when this brand oxidizes compounds such as
acetaldehyde, nonanol and furfural increase and other compounds such as myrcene,
and esters decrease. Some of the results indicate that the olive oil test was less
oxidized. For instance, acetaldehyde was 25% higher in the control beer than in the
test, nonanol was 10% higher in the control beer and most of the ethyl esters were
higher for the test brew. Some of the results indicate that the control was less
oxidized. For example furfural was 10% higher for the test beer and b-myrcene was
30% less in the test beer.
21
The sensory results for this round of testing also indicated very little difference
between the test and control. This round of testing was taken all the way through to
finished product and approximately 2100 hl of unblended test finished product were
packaged and released into the market by three separate taste screenings (finishing
taste release, packaging taste release and the brewery’s sensory panel).
The results of the sensory panel are given in Figure 7. Again the test was rated
higher in isoamyl acetate and sweetness but the differences were not statistically
significant. Some esters such as ethyl hexanoate and ethyl acetate were actually
perceived as being higher in the control. Based on the results shown in Figure 7, the
goal of this study, which was to achieve a flavor match without aerating the wort, was
achieved in this third test.
Figure 7. Sensory results for the third olive oil test vs. the control.
Sensory Test # 3 Fresh
control
olive oil test
hoppy
metallic
medicinal
diacetyl
fatty acid
oxidation
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
SO2
DMS
ethyl hexanoate
ethyl acetate
acetaldehyde
isoamyl acetate
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
One of the goals of this series of tests was to identify a process change that would
improve beer flavor stability without having a significant negative impact on finished
product quality or process efficiency. The results shown in Figure 8 do not show that
22
this process change significantly improves flavor stability. Although the control was
rated slightly higher in overall oxidation after 3 weeks at room temp, the difference is
obviously not significant.
The comments however, show that eight of the sixteen panelists said that the control
tasted oxidized but only four of them commented that the olive oil test was oxidized.
Other attributes remained approximately the same as the fresh test and were similar
to the control.
Figure 8. Sensory results for the third olive oil test after 3 weeks room temp.
Sensory Test # 3 at 3 Weeks Room Temp
hoppy
metallic
medicinal
diacetyl
fatty acid
oxidation
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
control
test
SO2
DMS
ethyl hexanoate
ethyl acetate
acetaldehyde
isoamyl acetate
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
Trial Four
In the fourth round of testing, again the GC results showed an increase in esters
such as ethyl hexanoate, ethyl acetate, and isoamyl acetate and a decrease in
acetaldehyde, DMS and higher alcohols. The results are shown in Figure 9. Based
on past results 2100 hl of this beer was packaged and sold without being blended.
23
The initial sensory results for this trial, shown in Figure 10, also indicated an increase
in ethyl hexanoate. The beer was released for sale by the in-house taste panels.
After three weeks warm storage the olive oil test was significantly less oxidized than
the control. It was also perceived by the panel as retaining more of the fresh beer
attributes such as ester and hop flavors (Figure 11). Overall the olive oil test was
preferred to the control.
Figure 9. Gas Chromatography results for the fourth olive oil test vs. the control.
GC Analysis Test # 4 Fresh
2.5
Flavor Units
2
1.5
Olive Oil Test
Control
1
0.5
an
oa
te
ol
lh
ex
al
et
hy
am
yl
is
o
am
yl
ac
et
ta
bu
is
o
is
o
co
h
at
e
l
no
ol
an
op
npr
la
ce
ta
t
e
dm
s
et
hy
ac
et
al
de
hy
de
0
24
Figure 10. Sensory results for the fourth olive oil test vs. control.
Sensory Test # 4 Fresh
metallic
medicinal
diacetyl
fatty acid
oxidation
hoppy
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
olive oil test
control
SO2
DMS
ethyl hexanoate
ethyl acetate
acetaldehyde
iso aa
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
Figure 11. Sensory results for the fourth olive oil test after 3 weeks warm storage.
Sensory Test # 4 at 3 Weeks Room Temp
metallic
medicinal
diacetyl
fatty acid
oxidation
hoppy
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
olive oil test
SO2
DMS
control
ethyl hexanoate
ethyl acetate
acetaldehyde
iso aa
bitter
H2S
sour
astringency
sweet/ malty
afterbitter
body
grainy
alcoholic
25
Discussion
This project set out to investigate the effects of using olive oil addition to storage
yeast vs. traditional wort aeration on a full production scale. Many similar studies
have been published using similar yeast treatments such as lineoleic acid addition to
yeast or yeast aeration (Fugiwara et. al., 2003, Depraetere et. al, 2003, Smart, 2000,
2003 Moonjai et. al., 2003a,b). The use of olive oil was selected for this study
because it contains oleic acid, which is the monounsaturated fatty acid naturally
produced by yeast cells.
The main areas of interest were fermentation performance, yeast health, ester
production and flavor stability. It was expected that if normal fermentations were
achieved flavor stability would be improved due to the reduced contact of oxygen
with the wort. It was also expected that if the olive oil treatment was unsuccessful,
as a means of supplying the necessary fatty acids, yeast growth would be reduced,
attenuation and VDK reduction would be incomplete, and that ester production would
be significantly increased.
Results
The results of this series of tests showed that normal production wort fermentations
could be carried out using yeast treated with olive oil instead of wort aeration. This
procedural change effected ester and fusel oil production, fermentation speed, and
overall flavor perception. Attenuation, pH, and foam were not affected. Ester
production was increased in all tests, although this increase was not deemed to be
out of specification for the brand by the flavor profile analysis panel. The rate of
attenuation in all trials was slower than the control samples but the fermentations
were complete and all final gravities were similar to controls. The overall effect on
oxidation potential in the final product was improved in the tests when compared to
the controls after three weeks warm storage. In the last round of testing the olive oil
test batch was judged by the panel to be significantly less oxidized than the control
after a period of warm storage. This method of treating the yeast with olive oil during
storage instead of aerating the wort did improve the overall flavor stability of the beer
without compromising flavor quality.
26
Oxidative Effects of Wort Aeration
Previously conducted research has shown that wort aeration causes oxidative
reactions to occur in the wort, which form the precursors to beer staling compounds
(Burkert, 2004). It is widely understood that minimizing the exposure of beer or wort
to oxygen will improve the finished product’s resistance to oxidation. During knock
out it is traditional practice to completely saturate the wort with air or oxygen,
intentionally dissolving 8 to 10 ppm oxygen into the liquid.
The yeast takes up the oxygen very quickly after wort aeration but the oxidative
reactions also take place very quickly. Even though wort aeration is a universally
excepted practice it seems very logical that eliminating this step would be a
significant improvement on the final beer’s resistance to oxidation. Therefore it
makes sense that reducing the product’s exposure to oxygen would improve its
flavor stability.
Ester Production and Growth
Another interesting aspect of this test was ester production. Low yeast growth has
been shown to cause a decrease in higher alcohols and an increase in esters
(Kunze, 1999). Oxygen is added to the wort so that the yeast cells can synthesize
sterols and fatty acids necessary for growth so it was expected that removing the
oxygen would cause an increase in esters and a decrease in higher alcohols. The
GC, and sensory panel results confirmed both of these outcomes. The
fermentations however were complete which indicates that the olive oil had a positive
effect on the fermentations. As the amount of olive oil was increased with each trial,
the fermentation performance improved. It is possible that the rate of fermentation
and the ratio of esters to higher alcohols could be improved if the amount of olive oil
addition were increased beyond the rate of 1 mg / 25 billion cells. For this brand, the
increase in total esters was perceived as preferable by the flavor panel.
Suggestions for Future Work
In order to achieve a healthy, vigorous fermentation yeast requires both sterols and
fatty acids. In this study only oleic acid was provided to the yeast. One suggestion
27
for future work would be to repeat this study adding a combination of ergosterol and
olive oil to the storage yeast. This would provide the yeast with both the unsaturated
fatty acid and the sterol normally produced by yeast in the presence of oxygen. It
would also be interesting to repeat this study adding only ergosterol to observe the
differences in its effect on attenuation, growth, esters, and flavor stability compared
to olive oil addition.
During storage yeast loses vitality and viability. It becomes vulnerable to conditions
that affect its metabolic activity. For this reason it has been proposed by Chris
Boulton (Smart, 2000) that any yeast aeration should be done at the start of yeast
storage rather than at the end. During storage yeast depletes its glycogen and
trehalose reserves, so oxygenating yeast after a prolonged period of storage can
trigger metabolic activity, which may be harmful to the yeast. Oxygenating yeast at
the beginning of yeast storage may provide the yeast with the sterols necessary for
proper fermentation at a time when the yeast is healthy enough to withstand some
metabolic activity (Smart 2000). This, together with olive oil addition, may provide
the combination of sterols and fatty acids necessary for a proper fermentation. The
addition of olive oil and sterols could also be combined with a reduction in wort
aeration to achieve a vigorous fermentation with reduced oxygen exposure.
Another suggestion for future work would be to continue with the addition of olive oil
to storage yeast but to experiment more with addition amounts and contact time.
Increasing the amount of oil addition, oil and yeast contact time, or yeast-pitching
rate may improve yeast health and growth, reducing ester formation and achieving
fermentations closer to the controls. Also other types of unsaturated fatty acids such
as linoleic acid could be used in place of olive oil. Fermentation parameters such as
temperature and pressure could also be manipulated to compensate for differences
in fermentation.
If the focus of the testing were taken away from flavor stability improvement, there
are a number of intriguing possibilities for future study. The addition of olive oil
during yeast storage, combined with wort aeration, could be used to test
improvements in fermentation performance. Alternatively the olive oil addition could
be tested in conjunction with zinc addition and wort aeration to optimize fermenter
turn around time and yeast health.
28
Conclusion
The idea of supplying yeast with the unsaturated fatty acids that it requires for
membrane health and growth is an intriguing possible alternative to the practice of
wort aeration. It has been proven that during the dormant phase of storage yeast
cells will take up fatty acids such as linoleic acid (Moonjai, 2003a). It is commonly
accepted that wort aeration is necessary for yeast growth but also that oxygenation
of the product reduces flavor stability. If the yeast is supplied with the olive oil during
storage, normal fermentations can be achieved with improved flavor stability. The
addition of olive oil to storage yeast in this study showed that consistent, complete
fermentations of acceptable flavor quality and improved flavor stability can be
achieved. However, these test fermentations were slower and produced an
increased amount of esters compared to the controls. Although the finished product
was significantly higher in ester content, it was not determined to be out of
specification and was actually preferred by the internal flavor panel. The goal of
improving flavor stability was achieved but at the cost of increased esters and slightly
slower fermentations.
29
Literature Cited
1. Anderson, R., Brites Sanches, A., Devreux, A., Due, J., Hammond, J., Martin,
T., Oliver-Daumen, B., Smith, I., (2000) Fermentation and Maturation.
European Brewery Convention Manual of Good Practice.
2. Burkert, J., Why Aerate Wort? A Critical Investigation of Processes During
Wort Aeration. Brauwelt International, 2004, vol. 22, 40-43.
3. Depraetere, S.A., Winderickx, J., Delvaux, F.R., Evaluation of the Oxygen
Requirement of Lager and Ale Yeast Strains by Preoxygenation. MBAA
Technical Quarterly, 2003, vol. 40, no. 4, 283-289.
4. Fugiwara, D., Tamai, Y., Aeration Prior to Pitching Increases Intracellular
Enzymatic and Transcriptional Responses Under Normal Non-nutritional
Conditions. Journal of the American Society of Brewing Chemists, 2003, vol.
61, no. 2, 99-104.
5. Hardwick, W. A., (1995) Handbook of Brewing.
6. Hough, J.S., Briggs, D.E., Stevens, R., Young, T.W., Malting and Brewing
Science Volume 2. 1982.
7. Kunze, W., (1999). Technology Brewing and Malting 2nd Ed.
8. Moonjai, N., (2003). Linoleic Acid Supplementation of Cropped Brewers
Yeast: Uptake and Incorporation into Cellular Lipids and its Effect on
Fermentation and Ester Synthesis. Dissertation, Katholieke Universiteit
Leuven
9. Moonjai, N., Verstrepen, K.J., Shen, H. -Y., Derdelinckx, G., Verachert, H.,
Delaux, F.R., Uptake of Linoleic Acid by Cropped Brewers Yeast and Its
Incorporation in Cellular Lipid Fractions. Journal of the American Society of
Brewing Chemists, 2003, vol. 61, no. 3, 161-168.
10. Smart, K., (2000) Brewing Yeast Fermentation Performance.
30
11. Smart, K., (2003) Brewing Yeast Fermentation Performance 2nd Ed.
31
Appendix 1
Fusel Oils by Headspace GC SOP
1.0
Purpose & Scope
This method separates and quantifies acetaldehyde, ethyl acetate, isobutyl
acetate, n-propanol, isobutanol, isoamyl acetate, n-butanol, isoamyl alcohol,
ethyl hexanoate, and dimethyl sulfide in cold wort, in process beers and
finished product.
2.0
Equipment & Chemical Preparation
•
Perkin Elmer Autosystem XL gas chromatograph
•
Perkin Elmer Turbomatrix headspace sampler
•
Perkin Elmer PE-Wax column (60m x 0.25mm id x 0.25um film)
•
20 ml Headspace vials and caps with septa and star springs
•
vial crimper
•
Ammonium Sulfate, granular (Mallinckrodt chemicals #7725)
•
n-propanol (Sigma Aldrich 29,328-8)
•
absolute alcohol (Sigma Aldrich
•
acetaldehyde
•
ethyl acetate
•
ethyl hexanoate (ethyl caproate)
•
isobutanol (2-methyl-1-propanol)
•
isobutyl acetate
•
isoamyl acetate
•
methyl acetate
•
2-methyl-1-butanol (active amyl alcohol)
•
3-methyl-1-butanol (isoamyl alcohol)
•
dimethyl sulfide
•
1-pentanol
•
pipettes, 250ul, 1 ml, 5 ml, and 10ml
•
Finn adjustable pipette
•
10 100ml volumetrics
32
•
1 liter volumetric flask
•
analytical scale
•
shaker table
3.0
Calibration
Stock Solutions
To prepare the stock solutions (FOA, FOB, FOC, and FOD) weigh the volume of
each compound given to 4 decimal places, into a 100 ml volumetric flask,
approximately half-full with absolute alcohol.
FOA
0.4 mls n-propanol
0.6 mls 2-methyl-1-butanol
0.6 mls 3-methyl-1-butanol
0.5 mls isobutanol
FOB
2.5 mls ethyl acetate
1 ml acetaldehyde (kept in lab fridge, freeze before attempting to pipette, and freeze
pipette)
0.1 ml isobutyl acetate
0.4 mls isoamyl acetate
FOC
0.1 ml ethyl hexanoate
FOD
0.1 ml DMS
33
Working solution FOE
Pipette 20 mls of FOA, 2mls of FOB, 1 ml of FOC, and 0.4 mls of FOD into a 100 ml
volumetric flask and make up to the mark with distilled water.
Standard solutions FO1, FO2, FO3, FO4, FO5, and FO6
Pipette the following volumes of FOE and absolute alcohol into separate 100ml
volumetrics, approximately half full with distilled water:
1 ml FOE, 3.8 mls absolute alcohol
2 mls FOE, 3.5 mls absolute alcohol
3 mls FOE, 3.3 mls absolute alcohol
4 mls FOE, 3.0 mls absolute alcohol
5 mls FOE, 2.8 mls absolute alcohol
6 mls FOE, 2.6 mls absolute alcohol
Make up to the mark with distilled water and label FO1, FO2, FO3, FO4, FO5, and
FO6 respectively.
Internal standard solution
Pipette 1 ml of methyl acetate and 2 mls of pentanol into a 1-liter volumetric flask
approximately half full with distilled water. Using a graduated cylinder, add 40 mls of
absolute alcohol to the flask and make up to the mark with distilled water. Decant the
solution into headspace vials (filling completely), crimp on the caps, and label FO Int
Std. Use a fresh vial with each analysis. You may decant some of the internal
standard into clean amber sample bottles, fill completely and label. Once this
container is open, you should decant the rest into headspace vials as described
above.
Calibration check solution
Pipette 14mls of absolute alcohol into a 500ml volumetric flask. Next pipette 25 mls
of FOE into the flask and make up to the mark with distilled water. Decant the
34
solution into clean, dry amber sample bottles (fill completely). Label the bottles as
FOCC.
4.0
Procedures
All samples, standards, and controls should be at or near 0C prior to preparation. In
process beers that are warm may be placed in the freezer for 10 minutes before
pipetting. Controls should be spaced between samples at a rate of approximately 1
control for every 6 samples.
Weigh 5.0 grams ± 0.1g of ammonium sulfate into a 20 ml headspace vial. It is best
to prepare around 30 vials at a time. Place the vials in the freezer for at least 30
minutes before using. Pipette 5 mls of sample or standard and 250ul of internal
standard into the vial and crimp on the cap. Always check the crimp for a good seal,
it should not spin easily. When all the vials have been prepped, transfer them to the
rotary shaker in the micro lab. Shake at 250 RPM for 30 minutes. During this time
you may log in the samples into the Turbochrome sequence. Run all samples at
least in duplicate.
Linear calibration
Perform the linear calibration every six months.
Prepare the standards FO1, FO2, FO3, FO4, FO5, FO6 in triplicate, as described
above. The Rsq should be above 0.98, if not the calibration should be repeated.
Calibration Check
Check the linear calibration by analyzing the standard calibration check solution
(FOCC) daily at the beginning and end of the run. Treat FOCC as a sample and
calculate the concentrations of the individual compounds. The calculated values
should fall in line with the known values ± 5%.
This files is hosted for you with Grady Hull’s permission by
www.BrewCrAzY.com
The Podcast for those of us who Brew our Own Damn BEER!
35